- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content
New vs. Old ESP Horsepower
10-25-2013
12:09 PM
Message 1 of 6
I was just playing around with ControlSpace Designer 4.0 to test DSP horsepower in the new ESP-Series models, and I’m seeing some very interesting results. This is far from an exhaustive test, just a quick peek.
Here’s what I did…
I have 4 different ESP units loaded into ControlSpace Designer:
A. Old ESP-88 with stock processing
B. Old ESP-88 with DSP expansion card
C. New ESP-00 II with 8x8 I/O
D. New ESP-880 stock
All four units are essentially 8 in, 8 out. The only exception is that the ESP-880 also has the digital ESP link card.
With those four units, I measured DSP horsepower with the following blocks:
1. Two 8x1 AutoMicMix blocks
2. Four 8-out Delay blocks
And here’s what I got…
Processors, in the order listed above (A through D) with the AutoMicMix blocks:
A. 93% DSP Usage
B. 30% DSP Usage
C. 30% DSP Usage
D. 21% DSP Usage (Whoa…what???)
Processors, in the order listed above (A through D) with the Delay blocks:
A. 74% DSP Usage, 14% Delay Usage
B. 22% DSP Usage, 4% Delay Usage
C. 22% DSP Usage, 4% Delay Usage
D. 10% DSP Usage, 38% Delay Usage
Couple-o-thoughts on this…
First off, the ESP-880 is going to be a beast of a DSP. Especially with fixed I/O, I don’t see horsepower ever being a real concern.
Second, this confirms what we’ve been told about the new ESP-00 II having the same horsepower as the old ESP-88 WITH the DSP expansion, for significantly less cost.
Third, it’s interesting to me that the ESP-00 II and the ESP-880 are so different in the way they process delay and DSP horsepower. Both are plenty powerful for most of our real-world applications, but the ESP-880 is a bit like cramming a Corvette motor into a Honda Civic. Do you need the power? Naw. But do you WANT the power? Yes, please. Very interesting...
Here’s what I did…
I have 4 different ESP units loaded into ControlSpace Designer:
A. Old ESP-88 with stock processing
B. Old ESP-88 with DSP expansion card
C. New ESP-00 II with 8x8 I/O
D. New ESP-880 stock
All four units are essentially 8 in, 8 out. The only exception is that the ESP-880 also has the digital ESP link card.
With those four units, I measured DSP horsepower with the following blocks:
1. Two 8x1 AutoMicMix blocks
2. Four 8-out Delay blocks
And here’s what I got…
Processors, in the order listed above (A through D) with the AutoMicMix blocks:
A. 93% DSP Usage
B. 30% DSP Usage
C. 30% DSP Usage
D. 21% DSP Usage (Whoa…what???)
Processors, in the order listed above (A through D) with the Delay blocks:
A. 74% DSP Usage, 14% Delay Usage
B. 22% DSP Usage, 4% Delay Usage
C. 22% DSP Usage, 4% Delay Usage
D. 10% DSP Usage, 38% Delay Usage
Couple-o-thoughts on this…
First off, the ESP-880 is going to be a beast of a DSP. Especially with fixed I/O, I don’t see horsepower ever being a real concern.
Second, this confirms what we’ve been told about the new ESP-00 II having the same horsepower as the old ESP-88 WITH the DSP expansion, for significantly less cost.
Third, it’s interesting to me that the ESP-00 II and the ESP-880 are so different in the way they process delay and DSP horsepower. Both are plenty powerful for most of our real-world applications, but the ESP-880 is a bit like cramming a Corvette motor into a Honda Civic. Do you need the power? Naw. But do you WANT the power? Yes, please. Very interesting...
Search by Labels: