Originally Posted by bobof
I don't think it is a matter of being lazy or not, just appropriate resolution. But if you've got the time it will be perhaps interesting. But I'm not sure how you can be so adamant of the importance given you haven't measured it yet! There is a lot of data for different PJs on Projection Dreams and a while ago I looked at the crossover points due to different on/off vs ANSI. As expected the on/off contrast differences massively dominates for a significant portion of the graph.
I'm pretty sure that the data to date show that the underlying relationships hold (assuming everything such as laser dimming, DI, etc is disabled).
I'm also pretty sure it could only ever be significant if there are two different models with really different ANSI contrast but almost identical on/off. In any case you'd know because they would have "crossed over" by 1% if the readings at the 0-1% end would be useful. Have you got any data yet for two different projectors where the lines have crossed over by 1%?
Here is the chart I did from the ProjectionDreams data:
Firstly, I am looking to take measurements that correspond to the contrast performance with actual usage; so that will most certainly include whatever dynamic contrast functionality a projector has (assuming it works to the extent that you would choose to use it). For example, are you going to watch movies via these new JVC projectors with the dynamic contrast turned off? No, you are not. That said, course I will also provide the native contrast measurements as well
Secondly, I want accurate graphs in their own right. I am measuring all kinds of different projectors with very different performances and technologies and I want 100% accurate contrast graphs, thanks. Where the more data points the greater the accuracy. For example, in those graphs you have posted there are a number of 'line of best fit' type estimates that I can see wherein the graphs might not be entirely accurate, especially between 0% and 1% ADL, as well as between 2% and 5% ADL given no measurements were taken between these data points. Which by the way is why we added 3% and 4% measurements; and the same applies to the range in-between 0% and 1% ADL.
For example, here's how that orange graph might actually track (IN RED)
Thirdly, why on earth are you arguing with me regarding my wanting to take just two extra measurements for increased accuracy?
And finally, here's an instance of measurements for you... please kindly tell me 100% accurately what the contrast measurement at 0.2% and 0.5% ADL will be?:
NATIVE ON/OFF CONTRAST (0% ADL) = 21,150,500:1
1% ADL = 83,649:1
2% ADL = 44,545:1
3% ADL = 34,730:1
4% ADL = 19,461:1
5% ADL = 15,452:1
10% ADL = 8,228:1
20% ADL = 3,771:1
ANSI CONTRAST (50% ADL) = 1,035:1