EDID: why a fuser has to look like a monitor
Every device in a video chain has to present an EDID block. Find out what that block holds, what a black screen or capped refresh rate means, and how EDID is stored, overridden and cloned.
Short answer
EDID is the block a display sends upstream listing its identity, supported timings and preferred mode. A fuser stands where the monitor used to, so it must present a valid EDID to each PC. A black screen or a capped refresh rate usually means that block is missing, generic or wrong.
On this page9
What is EDID?
Extended Display Identification Data is a small block of bytes that a display holds and hands upstream to whatever is driving it. The 128-byte base block carries the manufacturer and product identity, the EDID version, basic display parameters, colour characteristics, a legacy timing list, and up to four detailed timing descriptors. The first of those descriptors is the display's preferred mode. Later standards added 128-byte extension blocks, CTA-861 and DisplayID, which is where modern high-refresh modes, HDR metadata and audio capabilities live. A source reads this block, picks the best mode both ends support, and drives it; without a readable EDID, the source is guessing.
What does each EDID field control?
Each row gives a base block field and its practical effect, which is what to check when a chain lights up but runs at the wrong mode.
| Field | Where it sits | What it decides |
|---|---|---|
| Header | First 8 bytes, a fixed pattern | Whether the source believes it read an EDID at all. A bad header means the whole block is discarded |
| Manufacturer ID and product code | Bytes 8 to 11 | The name and model the operating system shows for the display |
| Serial number, week and year | Bytes 12 to 17 | Identity. Windows and GPU drivers key per-display settings off this, which is why a changed EDID can reset colour and scaling preferences |
| EDID version and revision | Bytes 18 and 19 | Which rules the rest of the block is parsed under |
| Video input definition | Byte 20 | Digital or analogue, and the digital interface and bit depth the display claims |
| Screen size or aspect ratio | Bytes 21 and 22 | Physical size. Windows uses it for DPI scaling recommendations |
| Gamma and feature support | Bytes 23 and 24 | Power management, colour type, and whether standard timing formulas may be used to derive modes |
| Chromaticity coordinates | Bytes 25 to 34 | Red, green, blue and white points. Wrong values here are the classic washed-out or oversaturated picture |
| Established and standard timings | Bytes 35 to 53 | The legacy mode list. This is where a fallback 1024x768 comes from |
| Detailed timing descriptors | Bytes 54 to 125, four 18-byte slots | The real modes, with pixel clock, blanking and sync. The FIRST descriptor is the preferred mode and is what a source picks by default |
| Extension block count | Byte 126 | How many 128-byte extension blocks follow. Zero here hides every mode those extensions would have described |
| Checksum | Byte 127 | Whether the block is trusted. A failed checksum is treated as no EDID at all |
Most support requests come down to two fields: the first detailed timing descriptor and the extension block count. A chain serving a valid but minimal EDID will light up at some safe low mode and then refuse everything the panel can really do.
Why does a fuser have to present an EDID?
A fuser is a sink on both of its inputs, so to each PC it is the monitor. It asserts hot plug detect on each input and then answers the DDC read with an EDID, and neither PC can see past it to the real panel. That gives the chain three EDID relationships rather than one: game PC to fuser, second PC to fuser, and fuser to monitor. The mode you end up with is the intersection of what the fuser advertises upstream and what the box can actually carry downstream.
Where is EDID stored, and how is it changed?
EDID lives in five different layers, and knowing which one you are working in saves a lot of wasted effort.
- In the display
- A small serial EEPROM read over the DDC lines, which are an I2C bus, at address 0x50. This is the source of truth for a real monitor
- In a pass-through device
- A fuser, splitter, switch or capture device stores an EDID of its own and serves that instead. Some can copy the downstream monitor's block, which is what a passthrough or clone mode does
- In the GPU driver
- Windows supports per-display EDID overrides in the registry. Tools such as ToastyX CRU write those overrides, changing what Windows believes without touching any hardware
- In a monitor INF
- The older mechanism for the same job: a driver package supplying an override block for one display
- In a dedicated emulator
- A small inline dongle that asserts hot plug detect and answers with a fixed EDID. Used in AV work to stop a source losing its display when a switch or KVM drops the connection
Why clone your monitor's own EDID?
A generic EDID is safe and small because it advertises modes almost anything can display. That becomes a problem when your panel is a 1440p 165 Hz display and the box in front of it advertises 1080p 60 Hz. Cloning copies the panel's real block, including its preferred timing, extension blocks and identity, so the game PC is offered the same mode list it would see with a direct cable. Clone rather than hand-write wherever you can, because a real block from a shipping monitor is internally consistent and a hand-edited one frequently is not.
How to clone or inject an EDID
The exact tool depends on the model, but the order of the steps does not.
- Cable the monitor directly to the game PC and confirm it runs the mode you want, because if it does not, the fuser is not your problem
- Read and save the monitor's EDID while it is directly connected, using the tool your vendor supplies or a standard EDID reader
- Cable the chain properly, with the game PC and the second PC in the box's inputs and the box's output going to the monitor
- If the model has a passthrough or clone mode, enable it first and re-test, since it is the lowest-risk option and is often enough
- If there is no such mode, or passthrough is not enough, write the saved block through the vendor's injection path. On several Dichen models, that path is a USB-C service port used only for firmware updates and EDID injection
- Power-cycle the box and re-plug the game PC's input so that hot plug detect fires and the EDID is read fresh
- Confirm the mode in Windows display settings, then set the refresh rate explicitly rather than trusting the default
- If you overrode EDID on the PC side with a registry tool instead, write that down, because the override survives cable changes and will confuse the next diagnosis
What each symptom means and how to fix it
Work from top to bottom, because the cheapest causes are listed first.
| Symptom | Likely cause | Fix |
|---|---|---|
| Black screen, monitor reports no signal | No valid EDID was read, or hot plug detect was never asserted, so the source has nothing to drive | Power the box before the PCs and re-seat both ends of every cable, using the bundled cables. Then enable the box's EDID passthrough or inject a saved block |
| Picture appears at a low resolution and will not change | The chain is advertising a minimal EDID, or the extension block count is zero so the high modes are invisible | Clone the monitor's native EDID into the box. If the box cannot hold one, its own ceiling may be the real limit |
| Correct resolution, refresh rate capped below the panel | Either the advertised EDID lacks that timing, or the box genuinely cannot carry the mode | Check the model's published ceiling at YOUR resolution first. No amount of EDID work raises a hardware ceiling |
| Game PC works, second PC gets no picture, or the reverse | The two inputs are separate EDID relationships and only one of them negotiated | Swap the two input cables to see whether the fault follows the cable or stays with the port, then treat that leg on its own |
| Windows shows a display named nothing like your monitor | You are seeing the box's own EDID identity rather than the panel's | This is expected on a fuser without passthrough, and it is only a problem if the mode list is also wrong |
| Colours wrong or washed out after inserting the box | Chromaticity or a colour-space claim in the served EDID does not match the panel, or an RGB and YCbCr range mismatch | Clone the panel's EDID, then check the GPU control panel's output colour format and dynamic range |
| Everything worked, then broke after a Windows or GPU driver update | A registry EDID override was cleared, or the driver renegotiated and chose a different mode | Re-check the override, set the refresh rate explicitly again, and power-cycle the box so the EDID is read fresh |
Common questions about EDID
Is EDID spoofing the same as spoofing hardware identity?
No, although the shared word causes real confusion. An EDID describes a display to a graphics output. It has nothing to do with a DMA card's firmware identity, a DNA ID, or a hardware spoofer for drive and board serials.
Can a wrong EDID damage a monitor?
Modern panels reject timings they cannot display, so the practical outcome is a black screen or an out-of-range message, not damage. Keep your changes reversible and keep a copy of the original block.
Do I need to clone EDID if my fuser has a passthrough mode?
Usually not. Try passthrough first, because it exists precisely so the box serves the panel's real capabilities with no manual work.
Why does a plain HDMI splitter cause the same symptoms?
The mechanism is the same: any device between the source and the panel becomes the thing answering the EDID read. That is why splitters, switches and KVMs produce identical mode and black-screen complaints.
Does changing EDID affect detection?
No. EDID is display negotiation between a graphics output and whatever sits downstream of it, not something an anti-cheat has a stake in. No EDID work makes a setup safe.
Sources
The fuser symptoms and the vendor injection steps come from bench work wiring these boxes, not from the EDID standard.
- 01VESA E-EDID standard, release A2
the 128-byte base block, the byte offsets in the field table and the detailed timing descriptors
glenwing.github.io
- 02EDID overview, Wikipedia
background on edid, its extension blocks and how a source reads it over the display's ddc lines
en.wikipedia.org
Read next
Still stuck? Open a ticket on Discord
All guides