DVD channel bit-rate question



Hello all and congrads on technical level of this forum,
My question is rather simple but I cannot find relative information somewhere on the web. Here it goes:

In order to calculate bit-rate for CDs (red book) you multiply the 588 bits frame size by 98 frames/sector and then by 75 sectors/sec and you get 4,321,800 bps, which is also mentioned on the red book. I am trying to find out where the 26.16 Mbps channel bit-rate of the DVD is coming from and I fail. Beginning with the 2064 byte sector through frame scrambling, ECC block & Recording frame formation, EFM+ modulation and down to recording frame I cannot correlate any structure size to match 26.16. It there someone with an answer??

Thank you,


It is pretty much the same for DVD as CD. It is easy to forget the SYNC bytes so maybe that is what you were missing.

38688 bits per sector after adding SYNC, PI and PO.
676.17866 sectors per second.

38688 * 676.17866 = 26159999.99808

1R (or 1X if you prefer) DVD has a 26.16MHz bit clock.


Thank you RichMan for your prompt reply. Actually I’ve tried Physical sector size also (38688 bits) but I was looking for an integer number of sectors/second. It is quite interesting that DVD was chosen to carry 676 sectors + an incomplete one/sec!!!


@RichMan Do you have any reference to this number: 676.17866 ?? (documentation)


Well data, I’ve often wondered that myself. In the end it doesn’t matter that there is a fractional sector per second but it is still very strange.

I do not recall ever seeing the sectors/second written in the specs. I think it is just derived from the bit clock and the channel bits/sector.

Here’s something to consider. I’ve heard that even though 1R DVD is referred to as 26.16MHz, the more accurate clock rate is 26,153,088Hz. Now if this is true, it makes the sectors/second an integral 676. But why would they have ever used the term 26.16Mhz in the first place? :confused:

Maybe there’s just not enough difference here to matter since the playback device just needs to supply the correct data rate to the decoder to achieve the proper video frames/second.

I’ll try to get the real answer once I return from vacation :slight_smile:


Thank you RichMan, in the mean time I will check 26,153,088Hz also. If something comes up I will post it here.


More info:

The DVD Physical Specification lists 26.16MHz as the nominal bit clock frequency at the nominal scanning velocity. It does not seem to specifically say anything about sectors/second.

The ECMA version of the spec lists the actual bit clock frequency (at reference scanning velocity) as 26.15625MHz. This would be 676.082 sectors per second.

There is about a 1% tolerance for scanning velocity. This allows for about 0.261 MHz deviation of the EFM+/bit clock signal.

It still seems odd to have a fractional sectors/second. But, I guess it really doesn’t seem to matter.



You are right RichMan. Probably this is tha major idea behind this…
You treat bit channel as a stream of bits like a telecom channel and you rebuild any logical stuctures after that.


There is no equivalence to calculate channel bit rate for DVD as you done for CD. For CD you should have said that one frame contains 6 audio stereo samples and so the sample rate is 75986= 44.1 khz the well know audio sampling frequency.
For DVD, the channel 1X bit rate has been choosen in order to provide around 11 Mbit/s of usefull data. This rate is the maximum rate allowed for a muxed (video, audio, subtitle) program in DVD-video. That’s all.


676.17866 sectors per second? Why not an even number?