• View Cart
    Shopping Basket CLOSE

    No products in the basket.

Blog CBHD Guides Open Questions
, , , , , , , , , , , , ,
CBHD Open Questions and Confirmed Facts

What the research passes treat as confirmed

After multiple independent compilation passes, the following statements are treated as established rather than speculative:

  • Professional CBHD authoring and professional CBHD verification existed (Memory-Tech/CBC/UOD production; DCA China HD verifier documentation).
  • UDF was used; CBHD had its own physical modulation (4-6 FSM replacing HD DVD 8-12 ETM).
  • Commercial CBHD players, proprietary optical pickups and CBHD-specific chipsets existed (Shinco and TCL production models; pickup and Novatek patents).
  • CBHD/BD optical compatibility was actively engineered (freeform objective-lens patent; multi-format pickup discussion).
  • Recordable CBHD was technically contemplated; commercial CBHD was primarily pressed/read-only (dye patent; AACS LA P/R categories; no shipped CBHD-R ecosystem found).
  • DKAA existed; AACS was used for international content; CETC navigation existed; AVS/DRA were part of the format.
  • CBHD was not normally compatible with HD DVD players.

Still unresolved – specifications and DRM

  1. Exact CBHD Book 2 and Book 3 specifications; complete CETC authoring specification.
  2. Complete DKAA specification, key architecture and key files. Expansions of the name disagree (“dual-key” vs “Data Key Authorization Access”); no pass cites a DKAA specification document; no DKAA patent numbers were located.
  3. GB/T, SJ/T or CY/T standard number for CBHD as a whole (AVS and DRA have numbers; the disc format does not in any pass).
  4. The actual FSM 4-6 patent (an OMNERC RLL patent exists but is not confirmed as commercial FSM).

Still unresolved – software and imaging

  1. Commercial authoring application names; encoder names and versions; CBHD muxer and premastering software (known to have lived at CBC; binaries not found).
  2. CBHD emulator; CBHD PC playback software (absence not fully proven).
  3. Firmware archive, firmware file formats and update history (THBD-1008 Spansion flash is a dump target only).
  4. An actual CBHD disc image; directory dumps; MediaInfo logs from real discs.
  5. Whether burned DVD/HD DVD-R media with a CBHD folder structure plays on real CBHD players (Gemini pipeline vs FSM physical argument).

Still unresolved – catalogue and hardware inventory

  1. Complete movie catalogue, catalogue numbers (ISRC/版号) and disc IDs.
  2. Complete player chipset inventory; service manuals; board photographs for every model; CBHD pickup part number.
  3. Commercial CBHD-R media and recorder.
  4. Any successful HD DVD → CBHD or CBHD → HD DVD conversion.
  5. Relationship between Shinco HD-3700 (HiFi demonstrator, July 2010) and HD-3700U (UTI hybrid player).
  6. Primary-source confirmation for TCL THBD-1038 feature set beyond secondary encyclopaedia entries.

Conflicts between passes that remain open

Topic Tension
PC HD DVD drives reading CBHD sectors Claimed workable vs FSM/channel hardware requirement; no published test
Home-burned DVD+R DL playback Gemini “works on HD-3700U/THBD-1008” vs no documented successful case elsewhere
Directory layout CBHD_TS/.AVS model vs EVOB/HD DVD Book 3 model
Early disc codecs “Exclusively AVS” vs early VC-1/AVC likely before Sep 2009
Region coding Undocumented / HD DVD-like free vs China-locked
Shinco monthly sales “Under 11,000” vs “on the order of 11,000” CHDA-sourced summaries
3:1 CBHD vs Blu-ray sales Single GOME store / promoter figure vs later CMM decline data

Highest-value next artefacts

  1. One inexpensive domestic CBHD disc plus sector image.
  2. Working Shinco CBHD-9100 and TCL THBD-1008 for behavioural comparison.
  3. Pickup label photos and Novatek/other chipset identification.
  4. NOR flash dump from THBD-1008 (and any Shinco dump).
  5. DCA Class A verifier with China HD configuration.
  6. CBC / Memory-Tech / UOD authoring or premaster samples.
  7. AACS CBHD Pre-recorded 0.90 book plus any DKAA fragment.
  8. std.samr.gov.cn / 新闻出版 search for CY/T CBHD documents.
  9. Full AVSForum master thread scrape while pages remain reachable.
  10. Wayback walks of shinco.com news/product trees and chdia.org.cn.

How to read the rest of this series against the open list

Format-at-a-glance and technical posts state physical parameters from the CHDA technical page with confidence. History and market posts carefully mark single-store sales myths. Codec posts separate what FFmpeg can decode from what DKAA still blocks. Hardware posts limit commercial models to the documented Shinco and TCL six, noting prototypes separately. Firmware and home-authoring posts present conflicting arguments side by side rather than picking a winner without evidence. Catalogue and buying posts refuse inflated 1,000-title claims and empty eBay searches as proof of total absence on Chinese C2C apps.

The open-question list is the research agenda. Confirmed items are the floor. Everything between-especially FSM vs burnable UDF, DKAA internals, and a real directory dump-decides whether CBHD reconstruction stays a literature project or becomes a playable archive.

Method notes for the next research cycle

Treat single-store sales myths, promoter-sourced 30/10/60 splits and alliance 1,000-title targets as claims with provenance, not as measurements. Treat CHDA physical parameters (0.204 μm pits, 0.40 μm track pitch, NA 0.65, 36.55 Mbps 1×, 28 Mbps max video) as higher-confidence because they come from the archived technical page. Treat Gemini-only pipelines and directory trees as hypotheses until a disc image confirms or falsifies them. Treat “no PC decoder found” as an open search, not a mathematical proof of absence.

When two passes conflict, publish both rows in the lab notebook rather than averaging them into a false consensus. FSM versus ATAPI sector-read, CBHD_TS versus EVOB, exclusive-AVS versus early-VC-1, and China-locked versus region-free are the four conflicts most likely to waste months if silently resolved the wrong way.

Success criteria for closing questions

  • Book 2/3: page images or PDFs with identifiable revision strings.
  • DKAA: patent numbers plus at least one key-file format description.
  • Authoring: named executable or installer from CBC/Memory-Tech with version.
  • Firmware: hashed NOR dump from THBD-1008 with readable strings list.
  • Disc image: full sector image plus UDF tree listing and MediaInfo.
  • Home burn: dated photo/video of a named burned disc playing on a named player model.
  • Catalogue: 版号 list or photographed spines covering >100 unique retail SKUs.
  • Pickup: part number photographed on a commercial Shinco or TCL mechanism.

Until those criteria trip, the series should keep hedging language. Confirmed facts in section one are the floor you can build tooling on; everything else is scaffolding. The corrected history-from 2005 WG-11 proposal through FSM/CETC/AVS engineering, Memory-Tech production, April 2009 launch, GBAX UK import, ~100-title plateau and 2011 collapse-is solid enough to motivate the hunt. The open list is what the hunt still owes the archive.

Hardware inventory gaps include service manuals for every commercial model, board photographs beyond the THBD-1008 excerpt, confirmation of THBD-1038 features from primary Chinese sources, and clarification whether Shinco’s HD-3700 HiFi demonstrator equals, variants, or differs from the HD-3700U UTI hybrid. Prototype photography of CBHD-9000 and Tongfang CBHD-0002 should stay labelled prototype, not quietly merged into the six-player commercial set.

Standards gaps include the missing whole-format GB/T or CY/T number, the true 4-6 FSM patent (as opposed to the nearby OMNERC RLL grant), and DKAA’s patent family. Software gaps include emulator existence, PC playback apps, and named muxer versions. Imaging gaps include any public sector dump. Compatibility gaps include any proven cross-flash and any proven burned-disc playback on named hardware. Closing any three of those clusters would materially change what “CBHD revival” can mean in practice.

Community-source gaps double as open questions: uncleared Tieba and Douban groups may hold directory screenshots; unrecovered GBAX product pages may hold UK bundle lists; unlocated BOC licensing pages may hold royalty schedules. The open list is therefore not only laboratory work-it is also archival web work against rotting 2009 HTML. Prioritise artefacts that unlock multiple questions at once: one disc image tests filesystem, codecs and DRM; one firmware dump tests CETC and key handling; one DCA verifier install tests Book 3 assumptions without needing the lost authoring GUI.

Ordering the hunt

If resources allow only three purchases, buy one domestic disc, one Shinco CBHD-9100, and one TCL THBD-1008. If resources allow only one engineering week, dump the THBD-1008 flash and photograph both pickups. If resources allow only web time, finish AVSForum page captures and std.samr.gov.cn searches for CY/T CBHD documents. Those three tracks-media, firmware, standards numbers-unlock the largest number of dependent questions on the unresolved list. Home-authoring experiments and cross-flash fantasies can wait until the floor artefacts exist.

Revisit this open list after each new artefact; strike items only when the success criteria above are met with citeable files.

Further reading on historic video formats

If you want to learn more about historic video formats, get The History of Physical Video Formats from the Frontier by Stephen Driver on Amazon:

Blog CVD Guides Open Questions
, , , , , , , , , , , , , , ,
CVD Open Questions and Confirmed Facts

Separating the known from the missing

China Video Disc (CVD) research is uneven: some commercial and technical facts are multiply attested, while neighbouring questions that collectors naturally ask-exact model lists, named titles, full standard texts-remain open. Separating the two is the point of this post. Confirmed items below are those the located 1998 Chinese press, C-Cube SEC filings, VCDImager documentation and DVD-Video references support. Unresolved items are framed as questions, not as guesses. Source conflicts that cannot yet be measured are kept as open disagreements rather than collapsed into a single “settled” story.

Confirmed facts

  1. Launch date and place. CVD was launched on 9 June 1998 in Shanghai with a jointly drafted industry standard (Wenzhai Bao).
  2. C-Cube decoder timing. C-Cube sold a CVD decoder from Q2 1998 and described it as Chao Ji VCD-compliant by Q3 1998 (SEC filings).
  3. National Super VCD standard. SJ/T 11196-1998 was announced on 29 September 1998, took effect on 1 November 1998, and made two technical changes towards CVD (Guangming Daily). What those two changes were is not yet recovered from the full text.
  4. Player mechanism and decoder camps. Contemporary Guangming Daily reporting associates CVD players with Philips 2× mechanisms, CD7 servo and C-Cube decoders, and SVCD players with Sony mechanisms and a reported ESS/Shinco decoder.
  5. 5.1-channel announcement. A 5.1-channel CVD scheme had been announced by November 1998 (Guangming Daily).
  6. Shared INFO system ID. CVD and SVCD share the “SUPERVCD” INFO system ID; video width (352 versus 480) is the reliable discriminator (VCDImager).
  7. Subtitle format. CVD-style subtitles are a distinct, non-IEC format, reverse-engineered in VCDImager’s cvd-subtitles documentation.
  8. MPEG-2 legality at 352 width. 352×480/576 MPEG-2 is DVD-legal video (DVD-Video / Wikipedia summaries of the DVD-Video constraints).

These eight points already constrain preservation work. A disc labelled only by the SUPERVCD system ID is not automatically “SVCD” in the 480-wide sense; width measurement is mandatory. Subtitle streams cannot be assumed to be standard OGT/IEC without inspection. Hardware identification should ask for Philips-versus-Sony mechanism photos and C-Cube-versus-ESS/Shinco chip markings because period reporting treated those as camp markers. C-Cube’s Q2-Q3 1998 SEC language ties the commercial chip story to Chao Ji VCD branding, which matters when Western sellers list “Chaoji” rather than “CVD.”

What “confirmed” does not mean

Confirmed does not mean complete. The 9 June Shanghai launch is attested; the full enterprise-standard text from that launch is not. SJ/T 11196’s effective date is attested; its full text and the exact “two technical changes towards CVD” are not. Mechanism/decoder camps are attested in contemporary comparison reporting; every individual 1998 retail SKU has not been inventoried. The 5.1-channel scheme was announced; shipping status of 5.1 CVD titles is unproven here. Shared SUPERVCD identity and 352-versus-480 discrimination are tool-supported facts; they do not invent a title list. DVD-legal video at 352 width explains why hobbyists could remux to DVD without re-encoding pictures; it does not prove every Western DVD player accepts CD-based CVD discs.

Still unresolved

The open list is long because the format’s paper trail is thin outside newspaper summaries and hobbyist reverse engineering:

  • Full text of the June 1998 CVD enterprise standard. Without it, “jointly drafted industry standard” remains a chronological fact rather than a readable technical specification.
  • Full text of SJ/T 11196-1998, and a precise account of the “two technical changes towards CVD” named in Guangming Daily.
  • Whether 1998 CVD allowed 480-wide video. Guangming Daily says CVD suited 2/3 D1 as well as 1/2 D1-language that would permit 480-wide as well as 352-wide material-but that claim sits in tension with later hobbyist practice that treats 352-wide as the CVD hallmark. Keep both notes; do not collapse them.
  • The exact CVD disc identifier and the mechanism of 1998 “interference codes” reported in the compatibility war.
  • Private_stream_1 sub-IDs for CVD-style subtitles. 0x00-0x03 is assumed in tooling discussion, not verified against a corpus of pressed discs.
  • Model names and chip markings of 1998 CVD players beyond the reported Jinzheng N100/N300. Brand-level coverage (Jinzheng, Aido, SAST and others) is far ahead of SKU-level coverage.
  • C-Cube CVD decoder part number(s) and datasheet. SEC filings confirm product timing and Chao Ji VCD compliance language; the discrete part numbers and electrical documentation remain missing.
  • Any named commercial CVD title, its publisher and 版号. Aggregate “several hundred titles” claims exist; individual spines do not, yet.
  • Identity of “Xiamin Solid” / “Xiamen Solid.” The name appears in C-Cube’s customer list and Western summaries as an unresolved entity label.
  • Whether current VCDImager warns or refuses on 352-wide SVCD input-a tooling behaviour question that affects how researchers classify homemade discs.
  • Real CVD compatibility rates of Western DVD players. Community claims range from roughly 30% of NTSC players to effectively 100% of SVCD players. Those figures disagree and must stay open.
  • Chipset support lists for later DVD platforms (Zoran, MediaTek, Sunplus) with respect to CVD playback.
  • Whether the “CVD 2000 project” (¥200 million) announced in November 1998 actually produced discs.

How to use the open list

Each unresolved item suggests a different next archive. Enterprise and SJ/T texts point at standards libraries, CNKI, and ministry or industry-body reprints. The 480-wide ambiguity points at recovering Guangming Daily’s full technical comparison table and measuring any surviving disc. Interference codes point at Wenzhai Bao’s late-1998 compatibility reporting and at physical discs that fail selectively on rival camps’ players. Subtitle sub-IDs require bitstream captures from labelled CVD discs, not more assumption. Model and chip work requires seller photographs, board shots and possibly Internet Archive captures of c-cube.com product pages from 1998-2000. Title and 版号 work requires Kongfz/Xianyu disc photography. “Xiamin/Xiamen Solid” needs Chinese-language company and press search against C-Cube’s named customer list. Compatibility percentages need structured playback tests across known SVCD-capable decks rather than another forum anecdote. CVD 2000 needs follow-up in late-1998 and 1999 industrial reporting.

Open disagreements to preserve

Two conflicts deserve explicit preservation as disagreements rather than premature synthesis. First, video width: tool culture and many Western guides treat 352-wide MPEG-2 as the CVD signature, while 1998 Guangming Daily language about suitability for 2/3 D1 as well as 1/2 D1 leaves room for 480-wide CVD. Second, Western player compatibility: one community tradition cites low NTSC success rates near 30%, another treats SVCD-capable players as effectively universal CVD players, and some VideoHelp posters reported never finding a working NTSC CVD machine at all. Until measured disc samples and controlled player matrices exist, both conflicts remain research status, not settled lore.

A third soft disagreement sits behind launch-maker rosters already discussed elsewhere in this series: Wenzhai Bao’s five-firm list, Jinzheng’s seven-firm “first batch,” Baidu Baike variants and Western summaries that name “Xiamen Solid” differently. Those are kept open in the history post; they reappear here only to note that “confirmed launch on 9 June 1998” does not yet mean “confirmed single maker list.”

A working research posture

The confirmed set is enough to begin serious collecting and dumping: know the launch date, know C-Cube’s 1998 decoder window, know SJ/T 11196’s calendar, know the Philips/C-Cube versus Sony/ESS-Shinco hardware axis reported in 1998, know that 5.1 CVD was announced, know that SUPERVCD identity is shared and width discriminates, know that CVD subtitles are a special case, and know that 352-wide MPEG-2 is DVD-legal video. The unresolved set is enough to stop false catalogue inflation: no standard PDFs in hand, no decoder datasheet, no named title, incomplete model list, unverified subtitle IDs, unsettled width and compatibility stories, and an unverified ¥200 million disc project.

Future posts in this series should promote an item from “unresolved” to “confirmed” only when a primary document, a measured disc, or a photographed SKU supports the move. Demoting a “confirmed” item should be equally explicit if a better primary source contradicts the newspaper or tool summary used here. CVD history will not be completed by smoothing over gaps; it will be completed by filling them with the same discipline that already secured the eight facts above.

Further reading on historic video formats

If you want to learn more about historic video formats, get The History of Physical Video Formats from the Frontier by Stephen Driver on Amazon:

Blog EVD Guides Open Questions
, , , , , , , , , , , , , , , ,
EVD Open Questions and Confirmed Facts

Separating the known from the missing

Enhanced Versatile Disc (EVD) research is uneven: some commercial and legal facts are multiply attested, while neighbouring questions that collectors naturally ask, exact model lists, named titles, container layout, codec of shipping discs, remain open. Separating the two is the point of this post. Confirmed items below are those the located English press, Chinese reference pages, E-World material and On2 SEC filings support. Unresolved items are framed as questions, not as guesses. Source conflicts that cannot yet be measured are kept as open disagreements rather than collapsed into a single settled story.

Confirmed facts

  1. Launch date and place. EVD was launched on 18 November 2003 at the Great Hall of the People with On2 VP5/VP6 announced and LSI Logic claiming the first HD EVD chips (EDN).
  2. On2 contract trail. The On2 contract terms, arbitration dismissal (10 March 2005) and unpaid 2006 settlement are documented in SEC filings.
  3. File-system standard entry. SJ/T 11299.2-2005 (file system) exists with 18 February 2005 dates, drafted by E-World and Shinco (standards platform).
  4. First player. The Shinco EVD-8700 was the first EVD player (1 January 2004, ¥1,998) and for a time the only one on sale (c114; Baidu).
  5. Named 2004 models. The Shinco EVD8830 and SVA EVD388 existed in 2004 (E-World material).
  6. December 2006 relaunch scale. The December 2006 relaunch showed about 54 players from about 20 makers at about ¥700 (AP).
  7. E-World software set. E-World had five software packages including an EVD disc authoring system and an OGT subtitle system (E-World material).

These seven points already constrain collecting. A deck without a period model story is not automatically “EVD” because a modern carton printed the three letters. The EVD-8700 / EVD8830 / EVD388 names are the documented early SKUs; the December 2006 “about 54 players” claim is a relaunch-era count, not a full serialised inventory. SJ/T 11299.2’s existence proves a file-system standard was filed; it does not put the full text of parts 1-4 into public hands. On2’s SEC trail proves contractual failure modes without proving which codec ultimately shipped on commercial discs.

What “confirmed” does not mean

Confirmed does not mean complete. The 18 November 2003 launch is attested; a title-level disc list from that day is not. VP5/VP6 was announced at launch; shipping-disc codec identity remains contested against later MPEG-2 MP@HL reporting. The EVD-8700’s first-player status is attested; the eleven launch players sometimes mentioned elsewhere and the fifty-four relaunch players are not inventoried by model here. E-World’s five software packages are named in secondary coverage; the binaries themselves are missing from public archives. SJ/T 11299.2’s catalogue dates are attested; whether the whole SJ/T 11299 family is still “current” or “withdrawn” is unresolved below.

Still unresolved

  • Which video codec shipping EVD discs actually use (MPEG-2 MP@HL-based vs VP6 vs other). Launch EDN coverage and later EE Times coverage disagree; keep both.
  • EVD directory and file names; container format.
  • Actual video bitrates of commercial discs.
  • ExAC bitrates, channel modes and whether it relates to Coding Technologies’ work.
  • The encryption scheme; whether any disc shipped with AACS.
  • Region coding, menus and navigation.
  • LSI Logic part numbers used in EVD players.
  • Model numbers of the 11 launch players and 54 relaunch players.
  • Any HD-EVD (2008) model.
  • A title-level disc list with 版号.
  • CNIPA patent numbers and the truth of the 2006 patent-pledge reports.
  • The “Hao Jie incident” of 2006.
  • Whether E-World ever paid the On2 settlement after 2009, and E-World’s corporate fate.
  • Current status of SJ/T 11299 (“current” vs “withdrawn”).

How to use the open list

Each unresolved item suggests a different next archive. Codec identity points at imaging a commercial disc and publishing ffprobe output after any decrypt path becomes available, and at re-reading EE Times 2006 beside EDN 2003 without forcing a premature synthesis. Directory and container work needs a published UDF file tree from a marked disc. ExAC questions need the SJ/T 11299.4 text or an E-World audio tool. Encryption and AACS questions need either the PC player binary (影海风雷 / E-World package) or a drive-authenticated read log. LSI Logic part numbers need board photographs from genuine players, not keyword-labelled portable DVD players from modern JD listings. Model inventories need 2004 and December 2006 retail photography and alliance press packs. HD-EVD needs 2008 product pages that name a chassis. Title and 版号 work needs Kongfz/Xianyu disc photography and national audio-video database queries. Patent-pledge claims need CNIPA record pulls. The Hao Jie incident needs the Huaqiang interview trail plus contemporary big-bit alliance reporting. On2 settlement and E-World fate need post-2009 corporate and SEC follow-up. SJ/T status needs a fresh standards-platform check against the existing 11299.2 catalogue entry.

Open disagreements to preserve

Two conflicts deserve explicit preservation as disagreements. First, video codec: November 2003 English launch coverage centres On2 VP5/VP6, while November 2006 EE Times relaunch coverage describes MPEG-2 MP@HL for commercial interest. Until measured streams from marked discs exist, both notes remain research status. Second, catalogue size: early counts of about 12-14 titles sit beside a Gome salesperson’s “hundreds,” E-World’s “more than 100,” and EE Times’ “about 100.” Those figures are kept side by side in the catalogue post; they reappear here only to warn that “confirmed launch” does not mean “confirmed filmography.”

A third soft disagreement sits behind player counts: “only EVD-8700 on sale” language in early reference pages, named EVD8830/EVD388 existence in 2004 E-World material, and AP’s December 2006 “about 54 players from about 20 makers.” Those are different moments, not a single contradictory census. Do not flatten them into one number.

A working research posture

The confirmed set is enough to begin serious collecting and dumping: know the launch date and On2/LSI Logic launch claims, know the On2 SEC contract trail, know that SJ/T 11299.2 exists on the standards platform, know the EVD-8700 first-player story and the EVD8830/EVD388 names, know the December 2006 relaunch scale language, and know that E-World shipped a professional software story including authoring and OGT tools. The unresolved set is enough to stop false catalogue and silicon inflation: no public container map, no measured bitrate table, no open ExAC decoder, no proven AACS-or-not answer, incomplete model lists, no HD-EVD SKU in hand, no title-level 版号 list, and unsettled patent-pledge and corporate-fate questions.

Modern shop keyword “EVD” still does not equal genuine 2004-2008 EVD with LSI Logic. Future posts in this series should promote an item from “unresolved” to “confirmed” only when a primary document, a measured disc, or a photographed SKU supports the move. Demoting a confirmed item should be equally explicit if a better primary source contradicts the press or FAQ summary used here. EVD history will not be completed by smoothing over gaps; it will be completed by filling them with the same discipline that already secured the seven facts above.

Further reading on historic video formats

If you want to learn more about historic video formats, get The History of Physical Video Formats from the Frontier by Stephen Driver on Amazon: